152 Mr. W. W. J. Nicol on 



of the other. My experiments were, I was fully aware, open 

 to the charge that they were made with equal volumes of solu- 

 tions of the same molecular constitution, and that therefore 

 the solutions mixed, though containing the same number of 

 salt molecules in proportion to the number of water molecules 

 present, did not contain an equal number of molecules of 

 either water or salt. In order to avoid this slight source of 

 error and at the same time to give my results a more definite 

 value, I have, in all my recent experiments, regarded not the 

 volumes of the solutions referred to water, but their molecular 

 volumes. Thus the quantities with which we have to deal 

 express the volume in cubic centimetres of the unit of the 

 solution expressed in grammes (1 gramme of water at the 

 given temperature being considered equal to 1 cub. centim.). 

 The volume is given by the formula 



Mol. vol.= s , 



o 



where #=the number of salt molecules (m) (in grammes) dis- 

 solved in 100 H 2 = 1800; 8= specific gravity of the solution. 



By mixing the solutions in the proportions of their mole- 

 cular weights (crm + 1800), mixtures were obtained which 

 should possess a molecular volume equal to the sum of the 

 molecular volumes of the two solutions before mixing. But 

 I have already shown (loc. cit.) that a slight change in volume 

 follows the mixture of salt solutions which cannot act on one 

 another, and that this change is a contraction. A slight dimi- 

 nution in the molecular volume is therefore to be regarded 

 merely as the result of the dilution of one solution by the 

 other. This change of volume never reaches '01 per cent., 

 even in solutions containing as many as 5 molecules of salt to 

 100 H 2 0. And even this slight source of error is eliminated 

 by the method of experiment. 



I have shown in a previous paper* that the volume of a 

 metal in a solution of its salt is independent of the salt radical 

 with which it is combined, and vice versa, if the solutions are 

 sufficiently dilute. Almost perfect density neutrality, as well 

 as thermal neutrality, exists in dilute solutions (less than *5 of 

 an equivalent of a salt in 100 H 2 0); but dealing with stronger 

 solutions, we find that density neutrality ceases to exist owing 

 to the varying solubility of the various salts that may be 

 formed, and that we have a tendency to form a system of salts 

 whose solutions have the smallest molecular volume; or if near 

 the saturation-point of any of the possible salts, that salt is 



* " Molecular Volumes of Salt-Solutions," Phil. Mag. August 1883. 



